size limitations (or what keeps us from building huge 3d printers)

[Tiago Santos]

Hey guys,

As I said on my previous post (intro and trades), I suffer from a bad case of scope creep. I made that post hoping to trade a bunch of parts and surplus for a smaller, functional printer or a complete kit that I can put together without much stress. But these things have a way of complicating themselves for me, which leads to this post. I have enough aluminum extrustions (45x90mm profiles) to build a pretty large printer (or actually 2 or 3! haha..).. say something like 20"x20". I also have a few linear motion bits - both 3/4" round rail stuff ("slightly" overkill..?) and THK 15mm square rails and bearing blocks (again.. slightly overkill, but they were fairly cheap).

I was wondering why no one seems to have/market/sell really large printers and figured there might be a reason for that. Are there complications in the actual printing process for large areas like that? Is it mostly the cost of materials (and the expense of oversizing all the components to deal with rigidity, inertia, etc..)? Or simply that no one seems to need large printed parts anyway?

To be fair, I don't think I need these large prints anyway. I'm mostly considering this as a way to put these parts to use. Something as big as what I'm thinking, and given that the linear motion stuff I have available is pretty hefty, would not be portable at all. I dunno, seems like a huge waste, which is why I initially thought it'd be great to trade with someone who actually needs this stuff for a router or plasma table or whatever.

Any thoughts are appreciated!

Cheers,
Tiago

[Loial Otter]

I'm making a large scale printer for sale; in this case large being around 1000x520x400mm. Due to plastic shrinkage factors this is about the maximum you can print PLA in that I've found. It even requires a certain ammount of trickery to get a reasonable print at 1m long. I'm working on methods which will make that less of a problem but don't have them built yet.

Speed is another factor, as you increase your print size the time it takes increases exponentially. I've been able to print up to 300mm/s though with some caveats. PLA limits speed more due to thermal mass (we need to increase our heater core wattage to make up for this) and takes more power in the drive motor.

If you want to see some info about my systems, i have a very out-of-date blog at ottersoft.ca which covers some of my tests and prototypes.

Let me know if you have interest in helping prototype my machines as well; I'm trying to get some aluminum milled for some of the components which has me stalled while i get funds together.

Regards,

    Loial

--
You received this message because you are subscribed to the Google Groups "3D604" group.
To unsubscribe from this group and stop receiving emails from it, send an email to 3d604+un...@googlegroups.com.
To post to this group, send email to 3d...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/3d604/8e5793df-804d-455e-b45a-05f88ec8bc0f%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[John Biehler]

Further to Loial's response, there are lots of reasons we don't see large FDM printers.

Cost is a big one. But plastic is always the main reason - most filament comes in 2-5lb spools so you'd have to roll your own larger spool for big job. Printing huge objects, even at 'fast' speeds would take days or even weeks unless you don't care about quality or resolution. Add to that the environmental issues with plastic printing (too hot, too cold, shrinkage, curling, cracking, drooping, bed unlevel, etc.) 

I know I'm not convinced that the current tech is good enough to work that long on a large object without a much higher failure rate...at least with plastic.

Use something other than plastic and it has a better chance of success but it's still pretty 'experimental'. 

There is no shortage of reasons to WANT a large printer though....print your own kayak, playground slide, Ironman suit, etc. but it's usually physics that stop it from being practical. That and it's usually much cheaper to make it another way at that scale.

To view this discussion on the web visit https://groups.google.com/d/msgid/3d604/CAHOeY1vvpeXCdsVUQiF7AMyW6hH8%3D89h29O6P1p-qLxKZaK3Lw%40mail.gmail.com.

[Kevin]

Re: Print head thermal mass issues   http://www.b3innovations.com/

To view this discussion on the web visit https://groups.google.com/d/msgid/3d604/CABSJ0FnwxtKUrwCTXJBj_sznC%3DuKpwZ5%2Bgf%3DFu7m-vN%3DCvWDNA%40mail.gmail.com.

[Kevin]

In general, the larger the axis are the heavier they get, then the speed vs accurately trade off becomes an issue.  You also have issues with bounce and vibration.

A massive printer is not an issue at all if your nozzle is proportionate to your build area.  Most printers have less than 200 square inches and their nozzle is 0.015 inches.  If you scale those dimensions together, you would end up with the same print speeds but larger objects.  Then the positioning accuracy can also be lower so it all works out.

[kburr]

Extrusions are great but they end up making a machine pretty heavy overall. So my suggestion would be for a stationary-doesn't-move-much machine would be just that. For portability, choose smaller extrustions or carefully plan it out. I say this as I built a MendelMax, and it is indeed a heavy beast to move around.

[Tiago Santos]

Loial, I'm more than happy to help with the machining stuff. We should discuss what you need off the board and if I can help with it, I will!

[Nick Wimpney]

well, generally you're laying down plastic on the edge of other plastic, which is often <.5mm wide, so you're going to need to have a machine rigid and accurate enough to hit that target.  Also, the prints do tend to warp a bit when they cool, even during printing, so there's limitations on how big you'll be able to print things anyway.

Also, there's the fact that what would be a minor defect in most machining operations will turn into a total failure on a 3d print.  The bigger the job, the more likely there is to be a problem.  and with bigger axes, any errors will typically be larger, and more likely to screw up the print.

I think time and expense are probably the biggest ones, though.  As it is, on a fairly fast printer, you're typically going to spend several hours printing an object. so double the dimensions and you're going to be measuring in days.

Also, plastic is typically >$20/kg for the cheapest stuff, and at some point it's going to make more sense to just switch to a different process and material, like milling, etc.

On Tue, Mar 25, 2014 at 11:58 AM, Tiago Santos <tiagosa...@gmail.com> wrote:

--

[Tiago Santos]

Thanks everyone, lots of great points to think about!

As far as the cooling, shrinking and warping issues, do enclosures have any sort of benefit? Again, is the lack of enclosures on most printers a cost saving thing or a practical consideration?

Thanks again :)

[Nick Wimpney]

No, the warping is just from the fact that the plastic contracts as it cools, so each layer gets printed imperceptibly bigger than the previous layer, and then shrinks slightly as it cools (pulling in at the edges of the previous layer that it's attached to).  After many more layers, these tiny stresses add up, and flex the object.

Controlling temperature with an enclosure can help some, but will never completely eliminate the problem.  With PLA, I don't think there's much that can be improved, since it doesn't properly harden until it's quite cool (<60C or so, at which point it's already shrunk?)  You generally want to cool the print as quickly as possible after it leaves the nozzle to give you a good surface to print the next layer on. 

With ABS, keeping the whole model quite warm will prevent it from shrinking (and warping) as much until it's finished printing, so the shrinkage should be more uniform, as opposed to flexing it on a single axis.  I haven't any experience with it either, but from what I've heard, cracking tends to be a problem in larger ABS prints, due to those stresses, unless there's lots of slots or holes to relieve them.

To view this discussion on the web visit https://groups.google.com/d/msgid/3d604/4ac653a2-7a83-48de-adf1-da95ba78d855%40googlegroups.com.